Multifaceted response curve from linear CV

[tysseng]

Hi!

First of all let me say that I'm an amateur when it comes to analog design, so don't be too hard on me

I am trying to create a circuit that transform a linear CV into a curve with several components. E.g, from 0 to 2V the input should be attenuated by a factor of 2, from 2 to 5 it should be attenuated by a factor of 1.5 or similar. The actual numbers are not important but the principle is.

See my attachments for a drawing of the response I want.

My approach to this is to make two separate amplifier parts. One that generates the first part (from 0 to 2v) and one that generates the second part, and thus is 0v untill the input voltage reaches 2V. I will then sum up the result to get the intended response

The first part is easy. For the second part, my idea is to create a circuit that attenuates the signal by the factor I want, then to add a negative voltage to offset the 0v crossing point so that V-out is 0v when V-in is 2v. Finally I add a diode to reject the negative part so that the response is 0 in the range 0 to 2v. (in my circuit diagram this diode does the oposite, it rejects positive voltages, since I am using an inverting summer).

The circuit I have drawn up works as intended (at fairly low frequencies, but that's ok, the input voltage is controlled using either an LFO or a pot, so high frequencies are not a problem.

What IS a problem however, is that when i try connecting this circuit to an additional summer, or even to a voltage buffer, it does not act very well.

How would I go about connecting this circuit to a summer - or if that is not possible, are there other ways of rejecting the negative/positive part of a signal without using a diode?
--
cheers,
Joakim

---

JT-response_curve.jpg
Description:	Wanted response curve, with my suggested approach of adding two separate curves.
Filesize:	742.25 KB
Viewed:	124 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

---

JT-circuit.jpg
Description:	My attempt at making a circuit that generates the second part of the response curve
Filesize:	266.27 KB
Viewed:	127 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[gdavis]

Your desired response looks like a linear approximation of an exponential. Does it really need to be two linear segments or could you use an exponential converter? Designs for the latter are pretty easy to find.

[Scott Stites]

This should get you where you want to be:

http://www.oldcrows.net/~patchell/archives/idealdiode.html

[Scott Stites]

For example, I was able to get this response out of a triangle wave (the triangle wave is bottom, the piecewise-linear derivation of the triangle wave is on top).

You can manipulate any CV to go either direction, at whatever gain or attenuation that you want, at any point you want by finagling these building blocks.

---

piecewise_linear_control.JPG
Description:
Filesize:	678.17 KB
Viewed:	125 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[elmegil]

You didn't make a batman logo, Scott!

[Scott Stites]

I'm Batman!

[tysseng]

Wow, thanks for the quick replies, I am stunned Scott: That was exactly what I was looking for. I tried it out, it works perfectly.

gdavis: It looks quite like an exponential function, and I did think about approximating part of it using an exponential function. The real curve goes much further to the right before it suddenly bends upwards. I did read up on antilogarithmic amplifiers using transistors, but was a bit set off by the thermal compensation problem. However, since you called it an exponential converter (I hadn't heard that term before, just antilog or exponential amplifier), I struck gold (for me) right after googling it, with MiK-Music's article https://plus.google.com/107231411209962227230/posts/C8iVtXKzvet

Thanks for all your help. Here is my first try:

---

P4154770.JPG
Description:
Filesize:	2.37 MB
Viewed:	128 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[gdavis]

Well I'd never seen the ideal diode circuit before, so I learned something too. That things pretty cool, I spiced it up and it really works great!

[Scott Stites]

Copyright that waveform ASAP. Then you can sell to AAA as a new, improved logo.

Bruce Wayne sent me a cease and desist.

[corex]

This is a very cool circuit... I should try it out. The possibilities are interesting.